When Difference between Kelvin and Celsius are negligible

[jldibble]

This question is more for practicality than anything else, and I realize it may be partly subjective.

At what temperatures can the 273 degree difference between Kelvin and Celsius scales be ignored? I'm thinking for examples regarding stars and very high temperatures. Surely once you get to 15 million it doesn't make much of a difference if you say the core of the Sun is 15 million K or 15 million degrees C.

I'm wondering what the lowest reasonable point is when the two units of temperature are basically interchangeable.

 

[hilbert2]

If you have something like a gas with a temperature of the order of 5000 Kelvin, and you want to apply the ideal gas equation of state to it, then the 273 K difference may not matter a lot. However, if there's a phase transition at some temperature, then a difference of 273 K is never insignificant.